ovs logs are trashed with healthcheck messages from ovslib

Bug #1687064 reported by Ihar Hrachyshka
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
neutron
Fix Released
Low
Unassigned
ovsdbapp
Invalid
Undecided
Unassigned

Bug Description

Those messages are all over the place:

2017-04-28 14:34:06.478 16259 DEBUG ovsdbapp.backend.ovs_idl.vlog [-] [POLLIN] on fd 14 __log_wakeup /usr/local/lib/python2.7/dist-packages/ovs/poller.py:246

We should probably suppress them, they don't seem to carry any value. If there is value in knowing when something stopped working, maybe consider erroring in this failure mode instead of logging in happy path.

Changed in neutron:
status: New → Confirmed
importance: Undecided → Low
tags: added: ovs usability
Revision history for this message
Terry Wilson (otherwiseguy) wrote :

We should not suppress them, they do have value. They are DEBUG logs and they are DEBUG information useful for finding race conditions. We specifically go to work to enable these logs--suppressing them would just be never setting them up in the first place with the vlog monkeypatching.

I agree with making them more targeted during testing, though.

tags: added: logging
Revision history for this message
Terry Wilson (otherwiseguy) wrote :

Fix proposed to branch: master
Review: https://review.openstack.org/562385

Changed in neutron:
assignee: nobody → Terry Wilson (otherwiseguy)
status: Confirmed → In Progress

Changed in neutron:
status: Confirmed → Fix Released
Changed in ovsdbapp:
status: New → Invalid
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.